IBM — International Business Machines Corporation

Sentiment is bruised-neutral, not uniformly bearish. The 2026-07-14 pre-announcement of an IBM Z shortfall and the subsequent cut to FY constant-currency revenue growth (4-5% from >5%) triggered a >25% single-session drop and $68B+ of lost market value on 2026-07-14, and the September 2026 Hagens Berman securities investigation keeps a legal overhang on the name. Major banks slashed targets hard (Morgan Stanley $293→$190, Susquehanna $303→$225, Citigroup $255→$245), yet as of 2026-09-13 the stock has clawed back +3.58% over the past week and +3.1% over the past month, and FMP's analyst breakdown (23 buy, 22 hold, 5 sell) shows the Street genuinely split rather than capitulating.

The revenue trend is a clear inflection down, not a steady deceleration. YoY revenue growth ran +1.5% (Q3 2024), +1.0% (Q4 2024), +0.6% (Q1 2025), then accelerated hard to +7.7% (Q2 2025), +9.1% (Q3 2025), +12.2% (Q4 2025) and +9.5% (Q1 2026) — before collapsing to +1.1% in Q2 2026 ($17.162B vs $16.977B). Operating margin followed the same arc: expansion through 2025 (Q3 2025 16.3% vs 14.5%; Q4 2025 24.6% vs 22.2%; Q1 2026 13.9% vs 12.1%) then a 300bp YoY compression in Q2 2026 (15.2% vs 18.2%), with diluted EPS flat at $2.30 vs $2.31. Crucially, cash flow did NOT break: Q2 2026 operating cash flow was $2.597B vs $1.701B a year earlier, and TTM free cash flow is $13.63B versus $11.81B for the prior TTM (+15.4%). Read: the deceleration is concentrated in the lumpy IBM Z hardware cycle, while the recurring software/consulting base and cash generation kept compounding — that pattern is consistent with a cyclical air pocket rather than broad structural deterioration, but one quarter is not proof.

Fair-to-mildly-expensive. As of 2026-09-13 IBM's TTM P/E of 21.25 looks cheap against its 5-year average of 35.58, but that historical average is distorted by depressed 2022-2024 earnings and the Q3 2024 net loss of -$0.33B. On the cleaner forward basis, $243.29 / $12.311 next-FY consensus EPS = 19.8x. Meanwhile price-to-sales of 3.317 is ~19% above the 5-year average of 2.79, EV/EBITDA is 17.9x, price-to-book is 6.6x, and the TTM free cash flow yield is only ~5.95% ($13.63B on a $229.2B cap). The headline PEG of 0.259 is misleading — it is anchored to trailing growth that is about to look much worse. Verdict: you are paying a full multiple for low-single-digit growth.

At $243.29, the market is pricing IBM at ~19.8x next-FY EPS of $12.31, which requires roughly 8% EPS growth on a revenue outlook management just cut to 4-5% constant currency. In other words, the shares have already de-risked the July shock but are not paying for a mainframe recovery either — a 'muddle through' price. The arbiter is the 2026-10-21 Q3 print: consensus is ~$2.90 EPS (non-GAAP; note FMP's figures are GAAP, so don't compare $2.90 to the $1.84 GAAP Q3 2025 number directly) on ~$16.88B revenue. Hit it with credible hardware commentary and the 200-day (~$261) comes into play; miss or guide down again and the multiple compresses toward the 17-18x the Street applied pre-2023, implying a retest of the low-$200s.

As of 2026-09-13, RSI(14) stood at 56.8 — neutral, neither overbought nor oversold. Price was 3.80% above the 20-day SMA (~$234) and 2.22% above the 50-day (~$238), but still 6.87% below the 200-day (~$261). That is a textbook post-crash recovery rally inside an intact longer-term downtrend, and the 20-day sitting below the 50-day says the short-term crossover has not fully healed. Entry-timing implication: momentum is improving but unconfirmed. The ~$261 (200-day) level is the line that flips the trend; failure there likely means a retest of the $199-210 zone. Chasing here is buying into overhead resistance, not a breakout.

• Enterprise IT budget reprioritization: clients are scrutinizing large hardware outlays, which is precisely what caused the July 2026 IBM Z shortfall, while generative-AI software and automation spend remains the protected part of the budget.
• Rates and leverage: with $56.2B of long-term debt against $7.2B of cash as of 2026-06-30, high borrowing costs pressure the Global Financing arm, raise refinancing costs, and elongate enterprise sales cycles.
• Regulation as tailwind: global data-sovereignty, cloud-security and emerging AI-governance mandates structurally favor entrenched hybrid-cloud vendors serving regulated banking, healthcare and government clients.
• Sector rotation: IBM straddles AI-momentum tech and defensive dividend/value, so it benefits from low-beta defensiveness (beta 0.708) but gets penalized every time hardware lumpiness resurfaces.
• FX / constant-currency framing: management now guides to 4-5% constant-currency revenue growth, so reported results carry translation risk stacked on top of an already-lowered bar.

• Q3 2026 earnings on 2026-10-21 — consensus ~$2.90 non-GAAP EPS on ~$16.88B revenue; the first real test of whether the z17 shortfall was a one-quarter air pocket. Hardware bookings and software growth commentary are the swing factors.
• Legal developments in the Hagens Berman securities investigation announced September 2026 — a formal complaint, class certification, or any hint of a restatement would be a fresh shock.
• Software mix proof point: Red Hat, automation and watsonx bookings and margin — the high-margin offset the entire bull case depends on.
• Mainframe cycle normalization: a strong Q4 2026 IBM Z quarter would retroactively label July as timing rather than demand destruction.
• Post-print analyst revisions and any acceleration of buybacks at depressed prices, which would mechanically lift EPS.

• A second guidance cut — hardware weakness proving structural rather than cyclical is the single biggest swing factor in the thesis.
• Securities class action: legal cost, discovery burden and management-credibility overhang; any accounting restatement would force a hard re-rate.
• Balance sheet: current ratio of 0.791 and debt-to-equity of 1.895 with $56.2B of long-term debt leaves a thin near-term liquidity cushion if rates stay elevated.
• Value-accretive growth gap: ROIC of 10.4% is low for the capital deployed, so incremental revenue growth may not translate into shareholder value.
• Competitive erosion: hyperscalers and global consultancies compete for the same AI/cloud transformation budgets, while mainframe workloads migrate to cloud over the long run.

Not a buy here. As of 2026-09-13 IBM is a levered, low-growth but genuinely cash-generative incumbent that just lost the Street's trust, and at ~19.8x forward EPS with a 5.95% FCF yield it screens as fairly valued — my blended fair value is roughly $235, about 3% below the $243.29 price, so there is no margin of safety even after a 27% drawdown from the high. The bull case (the IBM Z miss is cycle timing and software/AI keeps compounding) is plausible but unproven, and the 2026-10-21 Q3 print is the arbiter. Existing holders can justify holding for the 2.85% dividend and 0.708 beta; fresh money should wait for the print or a better price.

Blended two-method calc as of 2026-09-13: (1) a justified 19x forward P/E on next-FY consensus EPS of $12.311 = ~$234, below the 21.25x TTM P/E but above IBM's pre-2023 range given the improved software mix; and (2) a 6.0% required FCF yield on ~$13.4B normalized free cash flow = ~$223B market cap, or ~$236/share on 942M shares. The two methods converge near $235. The multiple reflects low-single-digit revenue growth, 10.4% ROIC, 1.895 debt-to-equity, a 2.85% dividend yield and a 0.708 beta — it is deliberately below the analyst consensus target of $256.69 because the class-action overhang and hardware-visibility gap warrant a discount.